Immobilisation of ketone catalyst: A method to prevent ketone catalyst from decomposing during dioxirane-mediated epoxidation of alkenes

Article abstract of DOI:10.1039/b005604i

Covalent attachment of trifluoromethyl ketone catalyst to a solid support such as silica gel increased the stability of catalyst dramatically, so solving an intrinsic problem of dioxirane-mediated epoxidation of alkenes catalysed by ketones.

Full text of DOI:10.1039/b005604i

Immobilisation of ketone catalyst: a method to prevent ketone catalyst from
decomposing during dioxirane-mediated epoxidation of alkenes
Choong Eui Song,*a Jin Seok Lim,b Su Chang Kim,b Kee-Jung Leeb and Dae Yoon Chic
a Life Sciences Division, Korea Institute of Science and Technology, PO Box 131, Cheongryang, Seoul, 130-650,
Korea. E-mail: s1673@kist.re.kr
b Division of Chemical Engineering and Industrial Chemistry, Hanyang University, Seoul, 133-791, Korea
c
Department of Chemistry, Inha University, 253 Yonghyun-Dong, Nam-Ku, Inchon, 402-751, Korea
Received (in Cambridge, UK) 11th July 2000, Accepted 30th October 2000
First published as an Advance Article on the web 21st November 2000
Covalent attachment of trifluoromethyl ketone catalyst to a
solid support such as silica gel increased the stability of
catalyst dramatically, so solving an intrinsic problem of
dioxirane-mediated epoxidation of alkenes catalysed by
ketones.
Dioxiranes have been shown to be powerful and environmen-
tally safe oxidants, which can be generated in situ from ketones
and Oxone® (2KHSO5·KHSO4·K2SO4).1 In principle, only a
catalytic amount of ketone is required and, moreover, with a
chiral ketone there exists the opportunity for catalytic asym-
metric epoxidation.2 However, an intrinsic problem of this
catalytic epoxidation procedure is that the turnover number
(TON) of ketone catalysts is generally very low owing to their
instability under the epoxidation conditions. It has been
generally known that ketone catalysts can be destroyed via
Baeyer–Villiger oxidation and/or formation of 1,2,4,5-tetrox-
ane etc.1
We proposed that the decomposition of ketone catalysts in
the dioxirane-mediated epoxidation of alkenes could be pre-
vented by covalent attachment of ketone catalysts to a suitable
solid support,3 since, as shown in Fig. 1, site isolation of
catalytic centres can minimise the possibility of 1,2,4,5-tetrox-
ane formation, and, moreover, the migration of a solid-bound
alkyl chain in the Baeyer–Villiger oxidation step may need
larger movement energy than that of the homogeneous
analogue. Of course, this approach can provide an additional
advantage of easy recycling of catalyst.
To test our rationale, we prepared a new silica-bound
trifluoromethyl ketone 1 and compared its stability in the
dioxirane-mediated epoxidation of alkenes with its homoge-
neous analogue, 1,1,1-trifluorododecan-2-one 2. Here we report
our preliminary results.
The silica-bound trifluoromethyl ketone 1 was simply
prepared by the reaction of 1,1,1-trifluorodode-11-cen-2-one 34
with 3-mercaptopropylsilanised silica gel 45 in the presence of
a,aA-azoisobutyronitrile (AIBN) as a radical initiator in acetoni-
trile (Scheme 1).6 The fluorine and sulfur analyses of 1 (F,
2.61%; S, 2.90%) indicated that 0.45 mmol g21 of tri-
fluoromethyl ketone moiety was incorporated and the coupling
yield of 3 with 4 was 50%. Thus, 0.45 mmol g21 of unreacted
thiol groups exist in the silica-bound trifluoromethyl ketone 1.
The attachment of trifluoromethyl ketone moieties to silica gel
in the silica-bound ketone 1 was also confirmed via IR
spectroscopy (1766 cm21 for nCNO).
In order to compare the catalytic efficiency and stability of
silica-bound ketone 1 with its homogeneous analogue 2, the
epoxidation of alkenes were carried out with Oxone® in a
MeCN–H2O solvent system under the modified condition
employed by Shi and co-workers.7 The results are summarised
in Table 1.
As shown in Table 1, the silica-bound ketone 1 exhibited
comparable catalytic activity with that of its homogeneous
analogue 2. In a ratio of ketone+olefin of 0.5+1 at room
temperature, epoxidation of alkenes using 2.07 equiv. of
Oxone® and 8.7 equiv. of K2CO3 was complete within 5 min
after addition of Oxone® and K2CO3, and afforded the
corresponding epoxides in very high yield.8 Most of the
oxidised products were obtained in nearly pure form by simple
filtration of the ketone 1, followed by extraction of organic
product and evaporation of the solvent. The observed results
suggest that trifluoromethyl ketone moiety on silica gel remains
highly exposed to the reactants and thus substrates can access to
the catalytic sites easily.
The silica gel supported trifluoromethyl ketone 1 could be
easily recovered after reaction by simple filtration and the
recovered ketone 1 was reused several times without any
significant loss of activity. All epoxidations of trans-b-
methylstyrene using the recycled ketone 1 afforded trans-b-
methylstyrene oxide in nearly quantitative yield, even after
tenth use.9 As we proposed, the retainment of the catalytic
activity during several cycles can be explained by that the
decomposition of the supported ketone catalyst 1 via Baeyer–
Villiger oxidation and/or 1,2,4,5-tetroxane formation10 pro-
ceeds much more slowly than that of its homogeneous analogue
2. We could not find any carbonyl peak of the Baeyer–Villiger
oxidation product (ester) in the IR spectrum of 1 recovered after
Fig. 1
Scheme 1
DOI: 10.1039/b005604i
Chem. Commun., 2000, 2415–2416
This journal is © The Royal Society of Chemistry 2000
2415

View Article Online
Table 1 Epoxidations of alkenes with Oxone® catalysed by silica-bound
ketone 1 or its homogeneous analogue 2a
ketone catalysts will provide a possibility to increase their
stability, which is an intrinsic problem of dioxirane-mediated
epoxidation catalysed by ketones. The current study will also
provide some insight for the development of new chiral ketone
catalysts. In the search for efficient ketone catalysts for
asymmetric epoxidation, both turnover number and enantiose-
lectivity are crucial issues to be considered.
This research was supported by a grant from Korea Institute
of Science and Technology.
Notes and references
1 R. W. Murray, Chem. Rev., 1989, 89, 1187; W. Adam, R. Curci and
J. O. Edward, Acc. Chem. Res., 1989, 22, 205; R. Curci, in Advances in
Oxygenated Process, ed. A. L. Baumstark, JAI, Greenwich CT, 1990,
vol. 2, ch. 1, p. 1; W. Adam, L. P. Hadjiarapoglou, R. Curci and R.
Mello, in Organic Peroxides, ed. W. Ando, John Wiley & Sons,
Chichester–New York–Brisbane–Toronto–Singapore, 1992, ch. 4, p.
195; R. Curci, A. Dinoi and M. F. Rubino, Pure Appl. Chem., 1995, 67,
811; W. Adam and A. K. Smerz, Bull. Soc. Chim. Belg., 1996, 105, 581;
W. Adam, A. K. Smerz and C.-G. Zhao, J. Prakt. Chem., 1997, 339,
298, and references therein.
2 Review: S. E. Denmark and Z. Wu, Synlett, 1999, 847.
3 Recently, simply for the easy separation of catalyst from the reaction
mixture, some polymer-bound ketones have been employed: T. R.
Boehlow, P. C. Buxton, E. L. Grocock, B. A. Marples and V. L.
Waddington, Tetrahedron Lett., 1998, 39, 1839; A. Shiney, P. K. Rajan
and K. Sreekumar, Polym. Int., 1996, 41, 377.
4 1,1,1-Trifluorododec-11-en-2-one 3 was synthesised by the published
procedure: J. Biovin, L. E. Kaim and S. Z. Zard, Tetrahedron, 1995, 51,
2573.
5 3-Mercaptopropylsilanised silica gel 4 (1.14 mmol of S g21) was
prepared by the reaction of Silica gel 60 (70–230 mesh, Merck) and
(3-mercaptopropyl)trimethoxysilane according to the published proce-
dure: C. Rosini, C. Bertucci, D. Pini, P. Altemura and P. Salvadori,
Tetrahedron Lett., 1985, 26, 3361; C. Rosini, P. Altemura, D. Pini, C.
Bertucci, G. Zullino and P. Salvadori, J. Chromatogr., 1985, 348, 79; P.
Savadori, C. Rosini, D. Pini, C. Bertucci, P. Altemura, G. Uccello-
Barretta and A. Raffaelli, Tetrahedron, 1987, 43, 4969.
6 Some of chiral catalysts have been anchored on silica by same
methodology: C. E. Song, J. W. Yang and H.-J. Ha, Tetrahedron:
Asymmetry, 1997, 8, 841; D. Pini, A. Mandoli, S. Orlandi and P.
Salvadori, Tetrahedron: Asymmetry, 1999, 10, 3883.
7 Z.-X. Wang, Y. Tu, M. Frohn and Y. Shi, J. Org. Chem., 1997, 62,
2328.
Scheme 2
8 The epoxidation of trans-stilbene in a MeCN–H2O solvent system under
the same reaction conditions7 reported by Shi (1.38 equiv. of Oxone®,
5.8 equiv. of K2CO3 and 0.5 equiv. of ketone) afforded the correspond-
ing epoxide in ca. 76% yield. However, the yield was dramatically
increased only by increasing the amount used of Oxone® and K2CO3. In
a 0.5+1 ketone : olefin ratio at room temperature, epoxidation of trans-
stilbene proceeded completely within 5 min after addition of 2.07 equiv.
of Oxone® and 8.7 equiv. of K2CO3 and afforded the corresponding
epoxide in nearly quantitative yield (Table 1).
9 There was some loss of ketone 1 owing to its slight solubility in the
reaction medium and thus, in a each run, the amounts of substrate were
used based on the amounts of the recovered ketone 1.
10 However, our site isolation hypothesis is not very convincing yet since
the loading of the ketone catalyst on the silica gel is so high that the
interaction between immobilised molecules can not be excluded.
11 Selected physical data of 6: 13C NMR (75.5 MHz, CDCl3) d 123.24 (q,
J 285.2 Hz), 94.69 (q, J 31.5 Hz), 72.09, 36.09, 32.15, 30.43, 29.64,
26.80, 25.57, 22.98, 18.18, 14.43; FAB-HRMS (M+ + Na). Calc. for
C24H42O4F6Na: m/z 531.2885. Found: 531.2886.
10 runs. In contrast to the silica-bound catalyst 1, ca. 85% of its
homogeneous analogue 2 was decomposed after a first run of
the epoxidation of trans-b-methylstyrene. The main decom-
posed products were ester 5 (ca. 8%) and 1,2,4,5-tetroxane 6
(ca. 77%) (Scheme 2). The structure of 6 was assigned by NMR
and HRMS analyses.11
In conclusion, we have achieved excellent results for the
heterogeneous dioxirane-mediated epoxidation of alkenes using
silica gel supported trifluoromethyl ketone 1. This silica-bound
ketone was reused ten times without loss of activity. This
retainment of the catalytic activity during several cycles was
explained by that the decomposition of the supported ketone
catalyst 1 via Baeyer–Villiger oxidation and/or 1,2,4,5-tetrox-
ane formation10 proceeds much more slowly than that of its
homogeneous analogue 2. Thus, this type of immobilisation of
2416
Chem. Commun., 2000, 2415–2416